This is a competing continuation proposal to extend results from a 5- year NIAAA-funded project on sertraline for depressed alcoholics. Comorbidity in substance abusers traditionally has been associated with a more severe clinical picture and a poorer prognosis for drinking outcome, compared to cases of uncomplicated alcoholism. In clinical populations, one-third to one-half of patients seeking alcohol treatment have a lifetime major depressive disorder. Persistent depression in abstinent alcoholics is both disabling and a risk factor for relapse to drinking, and further clinical deterioration that may result in suicide. Because we have effective, FDA-approved pharmacotherapy for alleviating depressive symptoms, it is important that we are fully informed about the advantages (or disadvantages) of treating primary or secondary depression in alcoholics with antidepressant medications. Results from our initial, project suggested that comorbidly depressed alcoholics appeared to have reduced antidepressant effects from sertraline and sertraline did not reduce their drinking (more than placebo). To address these results, we propose a study that will examine if we can achieve a more optimal outcome in comorbidly depressed alcoholics by directly treating the alcoholism with naltrexone, and combining this pharmacotherapy with the use of sertraline for treating depression. Thus, the primary aim of this proposal is to examine in depressed alcoholic outpatients whether combining naltrexone (an FDA-approved pharmacological intervention to reduce drinking) with sertraline (an FDA-approved pharmacological intervention to treat depression) will result in greater reductions in both drinking and depression over either medication alone or placebo. A secondary aim is to examine whether certain patient features, e.g., extent of pre-treatment drinking or severity of depression, will predict response to sertraline, naltrexone, or the combination. Patients who present to the University of Pennsylvania Treatment Research Center will be recruited for participation in this study over a 5-year period. There will be 160 males and females with a current DSM-IV diagnosis of alcohol dependence and also of major depression (via PRISM) who will be randomized to one of four treatment groups (40 per group): 1) the combination of 100mg/day naltrexone and 200mg/day sertraline, 2) 100mg/day naltrexone, 3) 200mg/day sertraline, or 4) placebo. Subjects will also receive once- weekly sessions of Cognitive Behavioral Therapy that has been adapted to include a medication compliance enhancement component. The treatment phase will last 16 weeks (includes a week of baseline, and a week of single-blind, placebo lead-in, and 14 weeks of double-blind pharmacotherapy). The follow-up phase includes two visits at 6 and 9 months post-treatment entry. Overall, this project will determine if combining pharmacotherapies results in a better response in comorbidly depressed alcoholics than either medication alone or placebo.